Download Journal of Biological Chemistry. 2004 Feb 20

JBC Papers in Press. Published on November 21, 2003 as Manuscript M307978200
The Kindler syndrome protein is regulated by TGFβ and involved in integrin-mediated
adhesion*
Susanne Kloeker, Michael B. Major, David A. Calderwood1, Mark H. Ginsberg1, David A.
Jones, and Mary C. Beckerle2
Department of Oncological Sciences, Department of Biology2, Huntsman Cancer Institute, Salt
Lake City, Utah 84112-5550
1
Department of Cell Biology, The Scripps Research Institute, La Jolla, California 92037
* This work was supported by a National Research Service Award (F32 20457 to S.K.), the
Multidisciplinary Cancer Research Training Program (T32 CA93247 to S.K.), the Huntsman
Cancer Foundation, the Willard L. Eccles Foundation, the National Institutes of Health Grant
GM50877 (to M.C.B.), the National Institutes of Health Grant HL48728 (to M.H.G.), the
American Heart Association Scientist Development Grant (to D.A.C.), and the University of
Utah DNA-Peptide Facility and Sequencing Facility Technical Support Grant CA42014.
2
To whom all correspondence should be addressed: Huntsman Cancer Institute, 2000 Circle of
Hope, University of Utah, Salt Lake City, UT 84112-5550. Tel.: 801-581-4485; Fax: 801-5812175; E-mail: [email protected].
1
Copyright 2003 by The American Society for Biochemistry and Molecular Biology, Inc.
Downloaded from www.jbc.org at University of North Carolina at Chapel Hill, on August 23, 2012
Running title: Kindlerin is a TGFβ-inducible focal adhesion protein
SUMMARY
TGFβ−1 contributes to tumor invasion and cancer progression by increasing the motility of
tumor cells. To identify genes involved in TGFβ-mediated cell migration, the transcriptional
profiles of human mammary epithelial cells (HMEC) treated with TGFβ were compared to
untreated cells by cDNA microarray analysis. One gene upregulated by TGFβ was recently
in this gene lead to Kindler syndrome, an autosomal-recessive genodermatosis (1,3). TGFβ
stimulation of HMEC resulted in a marked induction of kindlerin RNA and Western blotting
demonstrated a corresponding increase in protein abundance. Kindlerin displays a putative
FERM (four point one ezrin radixin moesin) domain that is closely related to the sequences in
talin that interact with integrin β subunit cytoplasmic domains. The critical residues in the talin
FERM domain that mediate integrin binding have been identified (4) and show a high degree of
conservation in kindlerin. Furthermore, kindlerin is recruited into a molecular complex with the
β1A and β3 integrin cytoplasmic domains. Consistent with these biochemical findings, kindlerin
is present at focal adhesions, sites of integrin-rich, membrane-substratum adhesion.
Additionally, kindlerin is required for normal cell spreading. Taken together, these data suggest
a role for kindlerin in mediating cell processes that depend on integrins.
2
Downloaded from www.jbc.org at University of North Carolina at Chapel Hill, on August 23, 2012
named kindlerin (1). This gene is significantly overexpressed in some cancers (2) and mutations
INTRODUCTION
The survival of cancer patients with solid tumors decreases dramatically when tumors are
invasive and have an increased likelihood of metastasizing to distal sites.
The enhanced
invasiveness of tumor cells is attributed to epithelial to mesenchymal transition (EMT), (5-8) a
normal biological process that is critical for wound healing and development.
EMT is
characterized by a change in cell shape from a polarized epithelial cell to a flattened fibroblast-
expression, and an increase in cell motility (5,6). Transforming growth factor β−1 (TGFβ)1 is a
major contributor to tumor progression and metastasis (9-11) and several studies have shown that
TGFβ promotes tumor cell invasiveness by promoting EMT (9,12-15). Despite the key role
established for TGFβ in stimulating EMT and tumor progression, the molecular mechanisms by
which TGFβ promotes EMT have not been fully elucidated.
Microarray analysis of a TGFβ-responsive cell line, human mammary epithelial cells
(HMEC), led us to identify kindlerin as a TGFβ-inducible gene. Kindlerin is mutated in Kindler
syndrome, a rare autosomal-recessive genodermatosis (1,3). Early in life patients with Kindler
syndrome endure blistering of the skin and photosensitivity, which progresses to diffuse
poikiloderma followed by cutaneous atrophy (16,17). The clinical presentation of this disease is
similar to patients with junctional epidermolysis bullosa harboring mutations in α6 and β4
integrin genes (18,19).
Kindlerin (also known as URP1 for UNC-112 related protein 1 or kindlin) is a member of
a newly recognized protein family, which also includes Mig-2 and URP2 (2,3). An apparent
kindlerin orthologue, UNC-112, has been studied in C. elegans (20). The unc-112 gene is
essential for embryogenesis and it displays a genetic interaction with integrins (20). Because
3
Downloaded from www.jbc.org at University of North Carolina at Chapel Hill, on August 23, 2012
like morphology, a decrease in cell-cell junctions concurrent with a decrease in E-cadherin
integrins play a critical role in mammalian cell adhesion and migration, we postulated that
kindlerin may be involved in TGFβ-stimulated EMT through interactions with integrins.
Here we report that kindlerin expression is responsive to TGFβ levels, that kindlerin
localizes in focal adhesions, (integrin-rich signaling centers that integrate extracellular matrix
attachment and cytoskeletal organization), and that kindlerin forms complexes with integrin β
Furthermore, cell spreading is perturbed upon reduction of
kindlerin protein. Taken together with the observation that kindlerin is overexpressed in colon
and lung carcinomas (2), our data support a role for kindlerin in mammalian cell adhesion and
suggest that kindlerin may mediate TGFβ signaling in tumor progression via contributions to
integrin-dependent cellular functions.
4
Downloaded from www.jbc.org at University of North Carolina at Chapel Hill, on August 23, 2012
subunit cytoplasmic domains.
EXPERIMENTAL PROCEDURES
Cell culture and drug treatments. HaCaT cells were maintained in DMEM supplemented with
10% fetal bovine serum and split every third day or at 80% confluency.
The HaCaT
immortalized keratinocyte cell line was a gift from D. Grossman. HMEC were obtained from
BioWhittaker (Maryland) and cultured in complete Mammary Epithelial Growth Media. HMEC
experiments.
cDNA microarray data analysis.
The construction of the microarrays, generation of the
microarray probes, microarray hybridization, and scanning were performed as described
previously (21). First-strand cDNAs were generated by reverse-transcription from the mRNA
samples in the presence of Cy-3dCTP or Cy-5dCTP.
The resulting labeled cDNAs were
combined and simultaneously hybridized to the microarray slide displaying 4608 randomly
selected and minimally redundant cDNAs from the Unigene set (22).
Each of the 4608
minimally redundant genes was present in duplicate on the microarray and the comparisons were
completed three separate times resulting in a total of six measurements for each gene. In each
case, mRNA from TGFβ-treated cells was directly compared to mRNA from vehicle-treated
cells. The GeneSpring software program (version 5.1; Silicon Genetics) was utilized for all steps
in the data analysis. To normalize the microarray data, a Lowess curve was fit to the log-intensity
versus log-ratio plot. 20.0% of the data was used to calculate the Lowess fit at each point. This
curve was used to adjust the control value for each measurement. We selected TGFβ responsive
genes based on a statistical analysis using the gene expression program GeneSpring (Version
5
Downloaded from www.jbc.org at University of North Carolina at Chapel Hill, on August 23, 2012
were seeded at passages 7 or 8 and harvested at no greater than 80% confluency for all
5.1). We applied the inclusion criteria of 1.5 fold induction and a p-value of less than 0.05
(Student’s T test) in defining TGFβ up-regulated genes. The list of genes shown in Table 1 were
selected from the total normalized data set as having a fold induction greater than 1.5 in at least 4
of 6 data points (see the experimental description within Table 1). Genes with a p-value
(Students T-test) greater than 0.05 were not included. DNA sequencing and BLAST analysis
confirmed the identity of each microarray clone corresponding to those genes listed in Table 1.
serum starved prior to the addition of growth factor. The vehicle control for TGFβ1 was
comprised of 4mM HCl, 1mg/ml BSA. Cyclohexamide (Calbiochem) was used at 10µg/ml and
was added to the cells 15 minutes prior to the addition of TGFβ. Total RNA was isolated using
Trizol (Invitrogen) followed by poly-A RNA selection using a PolyAT Tract mRNA Isolation kit
(Promega). Poly-A RNA was fractionated through formaldehyde-containing agarose gels and
transferred onto nylon membranes (Amersham Pharmacia). Probes were generated using the
Rediprime II random prime labeling system (Amersham Pharmacia) supplemented with
dCTP. Hybridizations with
32
P-
32
P-labeled probes were carried out using ULTRAhyb buffer
(Ambion) as recommended by the manufacturer.
The kindlerin probe was generated by
amplification of the insert of a cDNA clone (accession number 299593) using T7 and T3
primers.
Construction of Flag-tagged Kindlerin cDNA. In order to manufacture the longest piece of
the 5’ end of kindlerin cDNA, the EcoRI/SwaI fragment of AA158566 was subcloned into the
EcoRI/SwaI sites of AI147142 to make clone A. Clone A was used as a template for PCR to add
6
Downloaded from www.jbc.org at University of North Carolina at Chapel Hill, on August 23, 2012
Northern blotting. For treatments with TGF-β1 (Peprotech, New Jersey), HMEC cells were not
an EcoRI site directly 5’ of the ATG (sense oligonucleotide 5; 5’-ACG AAT TCA ATG CTG
TCA TCC ACT GAC TTT AC-3’) and ended at the native PstI site (antisense oligonucleotide 6;
5’-CGA GGA TGC TGC AGT TTT GTT CC-3’). The EcoRI/PstI insert from Clone A was
excised and replaced with the PCR product digested with EcoRI/PstI in order to remove the 5’
UTR to generate clone B. The 3’ end of kindlerin cDNA was isolated by RT-PCR using RNA
isolated from HMEC treated with TGFβ. The sense oligonucleotide 7 (5’-ATA GGT ACC TCA
(5’-GAA GCG GCG CTT TCT AAT TTG GA-3’) added a KpnI site directly 3’ of the stop
codon. The complete Flag-kindlerin cDNA was assembled by three-piece ligation containing
Clone B digested with EcoRI and HaeII, the RT-PCR product digested with HaeII and KpnI, and
FLAGpCMV2 digested with EcoRI and KpnI.
Kindlerin antibody generation.
A construct expressing the C-terminal third of kindlerin
(amino acid residues 500 to end) with a His-tag was engineered as follows: a PCR product was
generated using sense oligonucleotide 11 (5’-GCT CCA TAT GAT TCT TTC CTT TCT GAA
GAT GCA GCA T-3’) and antisense oligonucleotide 12 (5’-ATG CGG CCG CTC ACA CCC
AAC CAC TGG TAA GTT T-3’) using the FLAG-kindlerin cDNA as template and cloned into
NdeI and NotI sites of the pET28a vector (Novagen). The construct was verified by DNA
sequencing and utilized to transform competent BL21 cells (Novagen). The recombinant Histagged kindlerin fragment was purified under denaturing conditions in 6M urea. Fractions
containing purified protein were pooled and concentrated using a Centricon-30 (Millipore).
Concentrated protein was loaded onto SDS-PAGE, briefly stained with BioSafe Coomassie Blue
(BioRad), and excised from the gel. Gel pieces containing kindlerin were used to inject rabbits
for polyclonal antibody production (Harlan Bioproducts for Science).
7
Downloaded from www.jbc.org at University of North Carolina at Chapel Hill, on August 23, 2012
ATC CTG ACC GCC GGT CAA-3’) began at the HaeII site and the antisense oligonucleotide 8
Western blotting. For the TGFβ time course, HMEC or HaCaT were treated with vehicle or
TGFβ (2ng/ml) for the indicated times. Protein lysates were harvested from HaCaT or HMEC in
a buffer containing 25mM Tris-HCl pH 7.4, 150mM NaCl, 1mM CaCl2, 1% Triton X-100,
0.1mM phenylmethylsulfonyl fluoride, 0.1mM benzamidine, 1mg/ml Pepstatin A, and 1mg/ml
phenantholine. The lysates were centrifuged at 20,800x g for 10 min at 4oC. The supernatant
(BioRad). For antibody characterization (Figure 3A), lysates were heated at 70oC for 10 min in
LDS sample buffer (Invitrogen) and then separated by Tris-Glycine 10% SDS-PAGE and
transferred to nitrocellulose (Millipore). The blots were incubated with indicated kindlerin,
vinculin (Sigma), or E-cadherin (Transduction Lab) antibodies, followed by incubation with
donkey anti-rabbit horseradish peroxidase or sheep anti-mouse horseradish peroxidase
(Amersham). Immune complexes were visualized with Western Lightning Chemiluminescence
Reagent (PerkinElmer). For the time course experiments, lysates were harvested and analyzed as
indicated above, with the exception that Bis-Tris 4-12% NuPAGE using MES running buffer
(Invitrogen) was utilized to separate the lysates.
Phase contrast imaging. HMEC were allowed to adhere overnight and were subsequently
treated with vehicle or TGFβ for 48 hours. Cell morphology was observed using a Nikon
microscope and recorded with a digital camera.
Affinity purification of kindlerin antibody and indirect immunofluorescence. Kindlerin
antiserum was affinity purified using a method previously described (23). Briefly, antiserum
was incubated with recombinant kindlerin immobilized on nitrocellulose.
8
After extensive
Downloaded from www.jbc.org at University of North Carolina at Chapel Hill, on August 23, 2012
was transferred to a clean tube and assayed for protein concentration using the DC Protein Assay
washing of the membrane, the bound antibodies were eluted with 100 mM glycine, pH 2.5 and
immediately combined with 1M Tris-HCl, pH 9.0 sufficient to neutralize the glycine-HCl. The
eluted material was then concentrated in a Centricon-10 device (Amicon). The control antibody
was absorbed to nitrocellulose without kindlerin protein. Indirect immunofluorescence was
performed by plating HaCaT cells on glass coverslips coated with 10 µg/ml fibronectin (Sigma)
in Dulbecco’s modified eagle medium (Invitrogen) supplemented with 10% fetal bovine serum
vehicle or TGFβ (2ng/ml) for 48 hours. Coverslips were washed in PBS, fixed for 15 min with
4% formaldehyde, 5µM CaCl2, 0.5µM MgCl2 in PBS and permeabilized for 4 min with 0.2%
Triton X-100 in PBS. After washing in PBS, the coverslips were incubated with primary
antibodies at 37 oC for 120 minutes, washed in PBS for 10 min, and subsequently incubated with
Alexa-488 or Alexa-594 conjugated secondary antibodies (Molecular Probes). The coverslips
were mounted with ProLong (Molecular Probes) after washing for 10 min in PBS. Primary
antibodies used in this study include affinity purified rabbit polyclonal kindlerin antibody R2230 (1:100), control antibody (1:100), rabbit polyclonal talin antibody B-11 (1:800) and
vinculin monoclonal antibody (1:300 Sigma). Actin filaments were visualized using Alexa-488
conjugated phalloidin (1:100 Molecular Probes).
Cell images were visualized on a Zeiss
Axiophot microscope and recorded with a digital camera.
Kindlerin and talin amino acid alignment and modeling. The primary sequences of kindlerin
and talin were aligned by
BLAST
searching (http://www.ncbi.nlm.nih.gov/blast/) and refined
using structure based alignments performed by Swiss-PdbViewer (http://us.expasy.org/spdbv/)
using the structure of talin subdomains F2 and F3 (4)(PDB: 1MK7) as template. Based on the
9
Downloaded from www.jbc.org at University of North Carolina at Chapel Hill, on August 23, 2012
(Invitrogen). To examine kindlerin localization in the presence of TGFβ, cells were treated with
sequence alignment the putative kindlerin F3 subdomain (residues 566-655) was modeled on the
talin F3 subdomain (residues 309-400 from PDB 1MK7). Modeling was performed using Swiss
Model (24-26)and the quality of the resulting model assessed using WHAT_CHECK and
WHAT-IF version 19970813-1517 (27,28).
Integrin
tail binding assays. Affinity chromatography was performed using recombinant
with FLAG-kindlerinpCMV2 were lysed as described previously (29) and the indicated amount
of lysate mixed with 50 µl of His-Bind resin coated with recombinant integrin tails. Beads were
washed, bound proteins fractionated by SDS-PAGE and kindlerin detected by Western blotting
with FLAG antibodies. Loading of recombinant integrin tails onto the beads was assessed by
Coomassie Blue staining.
siRNA transfection. HaCaT cells were transfected using siPORT Amine (Ambion) with a 21nucleotide irrelevant RNA (control) or with a 21-nucleotide RNA targeting the kindlerin
sequence (only sense sequence is shown): 5’- GAA GUU ACU ACC AAA AGC UTT. Cells
were harvested approximately 44 hours after adding siRNA duplexes and examined for kindlerin
expression by Western blot analysis. To show equivalent loading of protein, the membrane was
probed with a talin monoclonal antibody (Sigma).
Cell spreading. Approximately 44 hours after addition of the siRNA duplexes, cells were
trypsinized and washed 2 times in OptiMEM (Invitrogen). Cells were plated on fibronectin (10
µg/mL, Sigma) coated coverslips and were allowed to spread for 30 minutes at 37oC/5%CO2.
10
Downloaded from www.jbc.org at University of North Carolina at Chapel Hill, on August 23, 2012
models of integrin cytoplasmic tails as described previously (29). Briefly, CHO cells transfected
Cells were stained by indirect immunofluorescence as described above using talin and vinculin
antibodies. A minimum of four randomly selected fields (approximately 180 cells total) was
recorded with a digital camera.
The area of cell spreading visualized by talin
immunofluorescence was measured using Openlab software.
Downloaded from www.jbc.org at University of North Carolina at Chapel Hill, on August 23, 2012
11
RESULTS
TGF treatment increases kindlerin expression. TGFβ elicits its biological effects by activation
of a SMAD-dependent transcriptional program (reviewed in (30-33)). To identify potential
downstream targets of TGFβ signaling that are involved in EMT and cell migration, we used
microarray analysis to compare the transcriptional profiles of human mammary epithelial cells
a greater than 1.5 fold induction and a p value of less than 0.05 (Table 1). Induced genes
included a number of known TGFβ responsive genes with roles in cell motility and adhesion
such as: plasminogen activator inhibitor type I, (34,35) transglutaminase, (36) thrombospondin 1
(37) and fibronectin (35). We examined Table 1 for novel genes with predicted roles in cell
adhesion and migration and recognized that the third ranking gene encoded kindlerin, a protein
required for stable attachment of epithelial cells to the lamina densa (1). Kindlerin displays
similarity throughout its sequence to two other human proteins, Mig-2 (65% identity) and URP2
(57% identity) as shown in Figure 1 and appears to be related to the C. elegans protein, UNC112 (44% identity). UNC-112 is required for the organization of dense bodies and attachment of
muscle cells to the hypodermis and UNC-112 interacts genetically with integrins (20). Thus we
focused on kindlerin as a possible mediator of TGFβ effects on integrin function.
In order to verify the induction of the kindlerin gene by TGFβ, mRNA was isolated every
2 hours up to 12 hours after treatment with TGFβ and examined by Northern blot analysis for
kindlerin expression (Figure 2A). A 4.6 kb kindlerin mRNA transcript was induced
approximately 10 fold 6 hours after addition of TGFβ. The level of kindlerin transcript remained
elevated after exposure to TGFβ for 96 hours (data not shown). To examine the role of protein
synthesis in regulation of kindlerin mRNA, we examined the effect of cyclohexamide on TGFβ
12
Downloaded from www.jbc.org at University of North Carolina at Chapel Hill, on August 23, 2012
(HMEC) treated with 10ng/ml TGFβ to those of vehicle treated cells. Seventeen genes exhibited
stimulated HMEC. Kindlerin mRNA abundance was markedly increased by the addition of
TGFβ (Figure 2B). Addition of cyclohexamide significantly reduced TGFβ’s effect on kindlerin
mRNA, indicating that protein synthesis is required for the TGFβ stimulation of kindlerin
transcription. Thus, kindlerin is a TGFβ-inducible gene.
We next assessed the effect of TGFβ on the abundance of the kindlerin protein. We
amino acids 500-677.
Western immunoblot analysis shows that the kindlerin antiserum
specifically recognizes a protein with an apparent molecular weight of 64,000 (Figure 3A) in
HMEC and HaCaT lysates. Because HaCaT and HMEC respond to TGFβ and express kindlerin,
these cell lines were used for our studies of the endogenous protein. The antiserum did not
recognize two kindlerin paralogues, Mig-2 and URP2, in mouse fibroblasts transfected with
FLAG-Mig-2 or FLAG-URP2 cDNA (data not shown); therefore, the antiserum is specific for
kindlerin.
To determine if TGFβ increases kindlerin protein levels, HMEC (Figure 3B) and HaCaT
cells (data not shown) were treated with TGFβ and lysates were prepared at various time points
up to 48 hours after initiation of treatment. Kindlerin levels increased slightly at 8 hours,
significantly at 24 hours, and most dramatically at 48 hours after TGFβ treatment. The induction
of kindlerin by TGFβ reached a plateau at 48 hours and remained elevated for up to 96 hours
(data not shown). The membrane was stripped and probed with vinculin antibodies to normalize
for equal protein levels. Kindlerin expression was quantitated relative to vinculin abundance at
the indicated time points (Figure 3C). Probing a parallel blot with E-cadherin antibodies showed
a decrease in expression upon TGFβ treatment, which is indicative of cells undergoing EMT.
Additionally, exposure of HMEC to TGFβ for 48 hours induces a change from a round, compact
13
Downloaded from www.jbc.org at University of North Carolina at Chapel Hill, on August 23, 2012
raised polyclonal antibodies directed against a bacterially expressed protein fragment containing
epithelial shape to a flattened fibroblast-like morphology (Figure 3D).
This change in
morphology is consistent with those observed for cells undergoing EMT in response to TGFβ.
These results demonstrate that kindlerin levels increase in the presence of TGFβ and indicate that
kindlerin may act as a downstream mediator of TGFβ-initiated EMT.
reorganization of the cytoskeleton (5,6). The kindlerin orthologue, UNC-112, co-localizes with
integrin adhesion receptors at dense bodies in the body wall muscle of C. elegans and is required
for the proper organization of integrin complexes at the site of muscle attachment to the
hypodermis (20). Dense bodies in worms and focal adhesions in mammalian cells share many
common components (38).
To determine if kindlerin is targeted to focal adhesions in
mammalian cells, immunofluorescence was performed using an affinity-purified kindlerin
antibody to detect endogenous kindlerin. As can be seen in Figure 4, kindlerin was concentrated
in focal adhesions where it colocalizes with vinculin. Vinculin and integrin adhesion receptors
are co-localized at focal adhesions in several cell types, including specialized epithelial cells
such as keratinocytes (39-43). Visualization of filamentous actin by fluorescent phalloidin
demonstrated that kindlerin is present at the tips of actin fibers (Figure 4D). No focal adhesion
staining was detected with a control antibody (Figure 4E) even though focal adhesions are
present and observed using a vinculin antibody (Figure 4F). Kindlerin was not detected at cellcell junctions that were compared by staining with E-cadherin antibodies (data not shown).
Thus, kindlerin is a constituent of focal adhesions.
14
Downloaded from www.jbc.org at University of North Carolina at Chapel Hill, on August 23, 2012
Kindlerin is a constituent of focal adhesions. EMT results in increased cell motility and a
Kindlerin localizes to membrane ruffles upon TGF treatment. Our data show that kindlerin is a
focal adhesion protein whose expression is regulated by TGFβ. In order to gain insight into the
role of kindlerin in TGFβ-mediated cytoskeletal reorganization, we examined its localization in
HaCaT cells exposed to vehicle (Figure 5A-B) or TGFβ (Figure 5C-D) and co-stained with
fluorescent phalloidin. After 48 hours of TGFβ stimulation a marked increase in cell spreading
vehicle treated cells, the actin cytoskeleton was organized as a network of filaments that
circumscribes each cell colony. After exposure to TGFβ, actin filaments were reorganized to
establish actin arrays that are typical of fibroblastic cells. A fraction of kindlerin remained in
focal adhesions and localized to the ends of actin fibers, but there was a notable increase in
kindlerin expression in a more diffuse cytoplasmic staining pattern. Co-staining HaCaT cells
with kindlerin and vinculin antibodies revealed a differential responsiveness of these two focal
adhesion constituents to TGFβ treatment (Figure 6). After TGFβ treatment, a significant loss of
coincidence of kindlerin and vinculin was observed in many but not all focal adhesions (Figure
6D-I). TGFβ resulted in an enrichment of kindlerin in membrane ruffles particularly at sites of
contact with other cells.
The focal adhesion constituent, vinculin, remained present at
extracellular matrix attachment sites indicating that focal adhesions had not been dissolved.
Kindlerin’s FERM domain is highly similar to the FERM domain of talin. As described above,
kindlerin localizes to focal adhesions and analysis of the kindlerin amino acid sequence suggests
a rationale for this localization. A search for conserved domains revealed that a striking feature
of kindlerin is homology of two regions of the protein with the FERM (four point one ezrin
radixin and moesin) domain (Figure 7A, gray, green, and blue boxes) and a predicted pleckstrin
15
Downloaded from www.jbc.org at University of North Carolina at Chapel Hill, on August 23, 2012
occurred, indicative of the transition from an epithelial to a mesenchymal cell morphology. In
homology domain (Figure 7A, black box). Interestingly, the pleckstrin homology domain sits in
the middle of a putative contiguous FERM domain and this architecture is a unique hallmark
shared by the members of the kindlerin family including Mig-2 and URP2.
Amongst other FERM domain proteins the kindlerin FERM domain has the highest
similarity to the FERM domain of talin, (44) an integrin binding protein that localizes to focal
adhesions (29). In order to gain insight into the function of kindlerin’s FERM domain, we
7B). The gray box at amino acid 250 in kindlerin (Figure 7A) illustrates where the similarity to
talin begins. FERM domains are comprised of three subdomains (F1, F2 and F3) arranged in a
clover leaf pattern (45) and often mediate interactions with the cytoplasmic domains of
transmembrane proteins. The regions similar to the F2 and F3 subdomains of talin are indicated
by green and blue, respectively. The structure of the F2 and F3 subdomains of talin has been
solved (4) and alignment of kindlerin and talin FERM domains predicts that the PH domain lies
in the large loop between the second and third alpha helix of the F2 subdomain. In talin, this
region does not comprise any of the core structure of the F2 domain. PH domains (Conserved
Domain Database: smart00233) can be inserted into other domains that retain activity
(http://www.ncbi.nlm.nih.gov/Structure/cdd/) and therefore, it is plausible that kindlerin houses a
functional PH domain within a structurally intact FERM domain.
The mutations identified in Kindler syndrome (as indicated by asterisks in Figure 7A),
yield truncated forms of kindlerin, all of which compromise the integrity of the FERM domain
(1). The longest disease-associated gene product is predicted to generate a protein identical to
kindlerin through the first 572 (arrow in Figure 7B) amino acids followed by three unique amino
16
Downloaded from www.jbc.org at University of North Carolina at Chapel Hill, on August 23, 2012
aligned the amino acid sequence of these corresponding regions in talin and kindlerin (Figure
acids (1). The resulting protein contains only five amino acids of the predicted F3 subdomain,
indicating the essential role of the F3 subdomain for kindlerin function.
The sequence similarities between F3 subdomains of talin and kindlerin permitted us to
build a reliable three-dimensional model of the kindlerin F3 subdomain (Figure 7C). As judged
by the program WHAT IF (27) the model has normal inside/outside distribution of
hydrophobic/hydrophilic residues (Z-score 1.005) and structural packing quality (Z-score -
Comparison of the model with talin F3 gives an rms of 1.32 A for 88 Cα atoms. Thus, the PTBlike (phosphotyrosine-binding) fold of F3 subdomains is compatible with the sequence of the
putative F3 subdomain of kindlerin.
The major integrin-binding site in talin lies within the F3 subdomain (46). Amino acids
participating in this interaction (4) are boxed in Figure 7B and the corresponding amino acids in
kindlerin are similar or identical. Modeling of the kindlerin F3 domain further highlights that
many of the residues conserved between talin and kindlerin, indicated by a teal color in the
kindlerin model, are clustered in the integrin binding region of the F3 subdomain (Figure 7C).
-integrin cytoplasmic domains form complexes with kindlerin. The amino acid conservation
between the talin and kindlerin FERM domains, particularly within the region responsible for
binding to integrin β tails, led us to hypothesize that kindlerin may bind to integrins. To test this
hypothesis, lysates from CHO cells expressing exogenous FLAG-kindlerin were incubated with
recombinant β-integrin tails (Figure 8A). Kindlerin was isolated with either the β1A or β3
integrin cytoplasmic domain. The association was specific as no kindlerin was isolated with the
cytoplasmic domain of integrin αIIb.
Tyrosine to alanine mutations within the conserved
17
Downloaded from www.jbc.org at University of North Carolina at Chapel Hill, on August 23, 2012
1.135), and the Ramachandran Z-score (-0.371) is within the range of well refined structures.
integrin NPXY motifs, that are known to disrupt a β turn (47) required for talin binding (46),
also significantly reduced kindlerin-integrin interactions. Furthermore, the binding of kindlerin
to the cytoplasmic domain of integrin β3 is dose-dependent (Figure 8B). Maximum binding was
observed with 50 µg of lysates. Hence, focal adhesion localization of kindlerin may be mediated
by FERM domain-integrin β tail interactions.
increase in integrin affinity for ligand (integrin activation) (46,48). The similarity of kindlerin to
talin’s FERM domain, in both amino acid sequence and ability to bind the cytoplasmic domain
of β integrins, therefore prompted us to examine the effect of overexpressed kindlerin on αIIbβ3
activation. Kindlerin expression had no effect on αIIbβ3 activation (data not shown). This
finding is similar to Numb and Dok1, other integrin β tail-binding proteins containing homology
to the F3 domain of talin, which are also unable to activate integrins (46).
Kindlerin is involved in cell spreading. The preceding data show that kindlerin is an integrinassociated focal adhesion protein. To assess the function of kindlerin, we employed small
interfering RNA (siRNA) duplexes to reduce the expression of kindlerin in human cells.
Kindlerin protein was reduced in lysates prepared from HaCaT cells transfected with kindlerin
siRNA as compared to lysates prepared from cells transfected with an irrelevant siRNA as shown
by Western blot (Figure 9A). The membrane was stripped and probed for talin to demonstrate
equal protein levels in both lanes. Forty-four hours after transfection of the siRNA duplexes,
cells were replated on fibronectin coated coverslips and allowed to spread for 30 minutes.
Indirect immunofluorescence analysis using talin antiserum revealed a decrease in the ability of
kindlerin siRNA cells to spread as compared to control cells (Figure 9B). To quantify the
18
Downloaded from www.jbc.org at University of North Carolina at Chapel Hill, on August 23, 2012
Binding of talin’s F3 subdomain to the cytoplasmic domains of β3 integrins leads to an
change in cell spreading, the cell area was measured for approximately 200 cells (Figure 9C).
The mean cell area was 712 µ2 for control siRNA cells and 452 µ2 for kindlerin siRNA cells.
Thus, reduction in kindlerin levels affects cell spreading.
Downloaded from www.jbc.org at University of North Carolina at Chapel Hill, on August 23, 2012
19
DISCUSSION
In this report, we have explored the molecular mechanisms underlying the cellular
response to TGFβ. We present data demonstrating that: i) kindlerin expression is increased
following stimulation with TGFβ, ii) kindlerin is a constituent of focal adhesions, iii) the integrin
binding residues found in talin are conserved in kindlerin, iv) kindlerin can be isolated in β-
We became interested in the kindlerin gene when examining our microarray data for
genes that may be involved in TGFβ-stimulated migration. Several pieces of circumstantial
evidence prompted us to investigate a possible role for kindlerin in mammalian cell adhesion and
migration. First, kindlerin contains two regions of homology with the FERM domain with a PH
domain inserted in between the two. The FERM domain has been identified in several molecules
such as ezrin, radixin, and moesin that connect the cytoplasmic domains of transmembrane
proteins to the actin cytoskeleton (44). Other proteins that display FERM domains are involved
in cell migration including focal adhesion kinase and talin (49). A second clue about the
potential function of kindlerin stems from studies of an apparent orthologue in C. elegans, UNC112.
UNC-112 null worms exhibit the pat (paralyzed arrested at two fold) phenotype, a
phenotype which is shared by α and β integrin, integrin linked kinase, and perlecan (20). GFPUNC-112 rescued embryonic lethality and colocalized with integrins at dense bodies in the body
wall muscle (20). In the absence of UNC-112, α-integrin failed to localize to the membrane in
body wall muscle (20). Moreover, UNC-112 was not properly targeted to the membrane in
worms lacking α-integrin implying an interdependence of UNC-112 and α-integrin for
organization of dense bodies (20). The unique domain architecture of kindlerin and the elegant
20
Downloaded from www.jbc.org at University of North Carolina at Chapel Hill, on August 23, 2012
integrin tail complexes, and v) kindlerin is required for normal cell spreading.
genetic studies of unc-112, led us to hypothesize that kindlerin is involved in mammalian cell
adhesion and migration.
We characterized kindlerin as a protein whose expression increases significantly in the
presence of TGFβ. Our studies indicate that enhanced transcription of the kindlerin gene is most
likely a secondary response to TGFβ stimulation because its induction is markedly reduced in the
mRNA and protein occurred at 6 and 48 hours, respectively, after treatment with the cytokine.
At 12 hours when the levels of kindlerin began to increase levels of E-cadherin, a marker for
EMT, began to decline.
Other reports have shown that EMT has taken place in HaCaT cells
after 48 hours of exposure to TGFβ (50) and the reduction of E-cadherin expression indicated
that our cells were responding in a similar fashion. The increase in kindlerin levels occurred
during the later stages of EMT progression and these data suggest that kindlerin may be involved
in the downstream events of the TGFβ response.
Indirect immunofluorescence analysis of kindlerin in HaCaT demonstrated that
endogenous kindlerin localizes to focal adhesions in mammalian cells. After TGFβ stimulation,
kindlerin is also detected at membrane ruffles particularly at sites of cell-cell contact. The
presence of kindlerin at focal adhesions is consistent with its isolation in integrin complexes and
implies a role in cell-matrix adhesion. A kindlerin-related protein, Mig-2, has also been shown
to localize to focal adhesions and is involved in regulating cell shape and cell spreading (51).
Mig-2 interacts with a protein termed migfilin, which potentially could connect Mig-2 to filamin
and the actin cytoskeleton. The presence of Mig-2 at focal adhesions is consistent with our
findings on kindlerin and supports a general role for the Mig-2 family in cell adhesion. To date,
no reports have described URP2 function; however, its high expression levels in spleen, thymus,
21
Downloaded from www.jbc.org at University of North Carolina at Chapel Hill, on August 23, 2012
presence of the protein synthesis inhibitor cyclohexamide. Additionally, peak levels of kindlerin
and peripheral blood leukocytes suggest a role specific to the immune system (2). Future
experiments will be needed to address the role of kindlerin’s enrichment at membrane ruffles and
sites of cell-cell contact in altered adhesion, migration and morphology of TGFβ-treated cells.
Based on the homology of kindlerin’s FERM domain with talin and the genetic
interaction in C. elegans between UNC-112 and integrins, we suspected that kindlerin interacts
In support of this hypothesis,
recombinant structural mimics of β1A and β3 integrin cytoplasmic domains were able to isolate
exogenously expressed FLAG-kindlerin. Although these experiments do not demonstrate a
direct interaction between kindlerin and β-integrin, similarity with talin and the absence of
kindlerin binding to the integrin β cytoplasmic domains containing tyrosine to alanine mutations
suggests that a direct interaction via the FERM domain of kindlerin is likely. We were unable to
confirm that the FERM domain was responsible for this interaction because of the insolubility of
carboxy-terminal kindlerin fragments (data not shown) and the low expression of amino-terminal
kindlerin fragments (data not shown).
Reduction of kindlerin by siRNA resulted in delayed cell spreading. Adhesion was not
affected by kindlerin knock-down when tested 30 minutes after replating on fibronectin or
laminin (data not shown). Therefore it is unlikely that kindlerin is required for integrin-ligand
interaction. The biochemical interaction between kindlerin and the cytoplasmic domains of β
integrins may contribute to efficient linkage of the extracellular matrix to the actin cytoskeleton
via integrin adhesion receptors. Moreover, the fact that kindlerin expression does not lead to
integrin activation suggests that the impact of kindlerin on cell spreading occurs after integrinligand binding. Whether kindlerin is affecting the formation of the actin cytoskeleton to enable
22
Downloaded from www.jbc.org at University of North Carolina at Chapel Hill, on August 23, 2012
with the cytoplasmic domain of the β-integrin subunit.
cell spreading or the assembly of the extracellular matrix to build focal adhesion attachment sites
remains to be elucidated.
The kindlerin gene was recognized by linkage analysis as being mutated in a rare skin
disease called Kindler syndrome.
This disease is characterized by acral blistering,
photosensitivity, and diffuse poikiloderma with cutaneous atrophy (1,3) and shares some
similarities in its clinical presentation with two subtypes of junctional epidermolysis bullosa
phenotypes have been described in mice with disrupted α3 (53) (a major integrin partner for β1
integrin in the basolateral membrane of keratinocytes), α6 (54), and β4 (55) integrin genes as
well as in mice genetically modified to specifically knock-out β1 integrin expression in skin
(56,57). Skin lesions of patients affected with Kindler syndrome have normal desmosomes,
hemidesmosomes, and anchoring filaments and the major defect lies in the organization of the
basement membrane which shows extensive reduplication and disruption of the lamina densa
(58,59). Interestingly, microscopic examination of skin from the α3 integrin null mice (60) and
the conditional β1 integrin null mice (56,57) also revealed a disorganized basement membrane.
The disruption of the integrity of the basement membrane caused by genetic defects in α3 and β1
integrin and kindlerin implicates these proteins in basement membrane assembly and
maintenance and suggests that they are likely components of the same signaling pathway. TGFβ
not only enhances kindlerin expression as shown in this report, but also increases the abundance
of integrins (61,62) and extracellular matrix constituents such as fibronectin, collagen, and
laminin (63-66).
Therefore, it is reasonable to postulate that organization of the newly
synthesized extracellular matrix components by integrin and kindlerin complexes may be an
important step in TGFβ-stimulated cell motility. All four of the mutations identified in Kindler
23
Downloaded from www.jbc.org at University of North Carolina at Chapel Hill, on August 23, 2012
which are caused by mutations in α6 or β4 integrin genes (18,19,52). Similar skin blistering
syndrome create truncated proteins that have lost the F3 subdomain of the FERM domain and
three of the four mutations generate products lacking both FERM domains and the PH domain.
This finding suggests that the F3 subdomain in kindlerin is necessary for stabilization of the
basement membrane through formation of integrin complexes.
Kindlerin has also been implicated in human cancer as it was found to be significantly
over-expressed in 60% of lung and 70% of colon carcinomas tested (2). Despite the similar
up-regulated (2). Interestingly, many colon cancers have increased levels of TGFβ expression
(67) which suggests that kindlerin is a TGFβ-inducible protein in vivo and evokes the possibility
that kindlerin is involved in tumor pathophysiology.
ACKNOWLEDGEMENTS
We acknowledge the expert graphics created by Diana Lim and Vera Tai for technical support.
24
Downloaded from www.jbc.org at University of North Carolina at Chapel Hill, on August 23, 2012
domain topology shared with Mig-2 and URP2, only kindlerin gene expression was found to be
REFERENCES
1.
Jobard, F., Bouadjar, B., Caux, F., Hadj-Rabia, S., Has, C., Matsuda, F., Weissenbach, J.,
Lathrop, M., Prud'homme, J. F., and Fischer, J. (2003) Hum. Mol. Genet. 12, 925-935
2.
Weinstein, E. J., Bourner, M., Head, R., Zakeri, H., Bauer, C., and Mazzarella, R. (2003)
Biochim. Biophys. Acta 1637, 207-216
3.
Siegel, D. H., Ashton, G. H., Penagos, H. G., Lee, J. V., Feiler, H. S., Wilhelmsen, K. C.,
Aboud, D., Al Aboud, K., Al Githami, A., Al Hawsawi, K., Al Ismaily, A., Al-Suwaid,
R., Atherton, D. J., Caputo, R., Fine, J. D., Frieden, I. J., Fuchs, E., Haber, R. M., Harada,
T., Kitajima, Y., Mallory, S. B., Ogawa, H., Sahin, S., Shimizu, H., Suga, Y., Tadini, G.,
Tsuchiya, K., Wiebe, C. B., Wojnarowska, F., Zaghloul, A. B., Hamada, T., Mallipeddi,
R., Eady, R. A., McLean, W. H., McGrath, J. A., and Epstein, E. H. (2003) Am. J. Hum.
Genet. 73, 174-187
4.
Garcia-Alvarez, B., de Pereda, J. M., Calderwood, D. A., Ulmer, T. S., Critchley, D.,
Campbell, I. D., Ginsberg, M. H., and Liddington, R. C. (2003) Mol. Cell 11, 49-58
5.
Savagner, P. (2001) BioEssays 23, 912-923
6.
Thiery, J. P. (2002) Nat. Rev. Cancer 2, 442-454
7.
Akhurst, R. J., and Derynck, R. (2001) Trends Cell Biol. 11, S44-51
8.
Derynck, R., Akhurst, R. J., and Balmain, A. (2001) Nat. Genet. 29, 117-129
9.
Oft, M., Heider, K. H., and Beug, H. (1998) Curr. Biol. 8, 1243-1252
10.
McEarchern, J. A., Kobie, J. J., Mack, V., Wu, R. S., Meade-Tollin, L., Arteaga, C. L.,
Dumont, N., Besselsen, D., Seftor, E., Hendrix, M. J., Katsanis, E., and Akporiaye, E. T.
(2001) Int. J. Cancer 91, 76-82
25
Downloaded from www.jbc.org at University of North Carolina at Chapel Hill, on August 23, 2012
South, A. P., Smith, F. J., Prescott, A. R., Wessagowit, V., Oyama, N., Akiyama, M., Al
11.
Janda, E., Lehmann, K., Killisch, I., Jechlinger, M., Herzig, M., Downward, J., Beug, H.,
and Grunert, S. (2002) J. Cell Biol. 156, 299-314
12.
Boland, S., Boisvieux-Ulrich, E., Houcine, O., Baeza-Squiban, A., Pouchelet, M.,
Schoevaert, D., and Marano, F. (1996) J. Cell Sci. 109, 2207-2219
13.
Bakin, A. V., Tomlinson, A. K., Bhowmick, N. A., Moses, H. L., and Arteaga, C. L.
(2000) J. Biol. Chem. 275, 36803-36810
Ellenrieder, V., Hendler, S. F., Boeck, W., Seufferlein, T., Menke, A., Ruhland, C.,
Adler, G., and Gress, T. M. (2001) Cancer Res. 61, 4222-4228
15.
Bhowmick, N. A., Ghiassi, M., Bakin, A., Aakre, M., Lundquist, C. A., Engel, M. E.,
Arteaga, C. L., and Moses, H. L. (2001) Mol. Biol. Cell 12, 27-36
16.
Kindler, T. (1954) Br. J. Dermatol. 66, 104-111
17.
Forman, A. B., Prendiville, J. S., Esterly, N. B., Hebert, A. A., Duvic, M., Horiguchi, Y.,
and Fine, J. D. (1989) Pediatr. Dermatol. 6, 91-101
18.
Vidal, F., Aberdam, D., Miquel, C., Christiano, A. M., Pulkkinen, L., Uitto, J., Ortonne,
J. P., and Meneguzzi, G. (1995) Nat. Genet. 10, 229-234
19.
Pulkkinen, L., Kimonis, V. E., Xu, Y., Spanou, E. N., McLean, W. H., and Uitto, J.
(1997) Hum. Mol. Genet. 6, 669-674
20.
Rogalski, T. M., Mullen, G. P., Gilbert, M. M., Williams, B. D., and Moerman, D. G.
(2000) J. Cell Biol. 150, 253-264
21.
Karpf, A. R., Peterson, P. W., Rawlins, J. T., Dalley, B. K., Yang, Q., Albertsen, H., and
Jones, D. A. (1999) Proc. Natl. Acad. Sci. U. S. A. 96, 14007-14012
22.
Miller, G. S., and Fuchs, R. (1997) Comput. Appl. Biosci. 13, 81-87
23.
Beckerle, M. C. (1986) J. Cell Biol. 103, 1679-1687
26
Downloaded from www.jbc.org at University of North Carolina at Chapel Hill, on August 23, 2012
14.
24.
Guex, N., and Peitsch, M. C. (1997) Electrophoresis 18, 2714-2723
25.
Peitsch, M. C. (1996) Biochem. Soc. Trans. 24, 274-279
26.
Peitsch, M. C. (1995) Biotechnology 13, 658-660
27.
Vriend, G. (1990) J. Mol. Graph. 8, 52-56, 29
28.
Hooft, R. W., Vriend, G., Sander, C., and Abola, E. E. (1996) Nature 381, 272
29.
Calderwood, D. A., Zent, R., Grant, R., Rees, D. J., Hynes, R. O., and Ginsberg, M. H.
30.
Massague, J. (1998) Annu. Rev. Biochem. 67, 753-791
31.
Christian, J. L., and Nakayama, T. (1999) BioEssays 21, 382-390
32.
Verrecchia, F., and Mauviel, A. (2002) J. Invest. Dermatol. 118, 211-215
33.
Derynck, R., and Zhang, Y. E. (2003) Nature 425, 577-584
34.
Gerwin, B. I., Keski-Oja, J., Seddon, M., Lechner, J. F., and Harris, C. C. (1990) Am. J.
Physiol. 259, L262-L269
35.
Stampfer, M. R., Yaswen, P., Alhadeff, M., and Hosoda, J. (1993) J. Cell. Physiol. 155,
210-221
36.
Akimov, S. S., and Belkin, A. M. (2001) J. Cell Sci. 114, 2989-3000
37.
Claisse, D., Martiny, I., Chaqour, B., Wegrowski, Y., Petitfrere, E., Schneider, C., Haye,
B., and Bellon, G. (1999) J. Cell Sci. 112, 1405-1416
38.
Burridge, K., and Chrzanowska-Wodnicka, M. (1996) Annu. Rev. Cell Dev. Biol. 12,
463-518
39.
Kramer, R. H., McDonald, K. A., and Vu, M. P. (1989) J. Biol. Chem. 264, 15642-15649
40.
Rahilly, M. A., Salter, D. M., and Fleming, S. (1991) J Pathol 165, 163-171
41.
Gates, R. E., Hanks, S. K., and King, L. E., Jr. (1993) Biochem. J. 289, 221-226
27
Downloaded from www.jbc.org at University of North Carolina at Chapel Hill, on August 23, 2012
(1999) J. Biol. Chem. 274, 28071-28074
42.
Horiba, K., and Fukuda, Y. (1994) Virchows Arch 425, 425-434
43.
Levy, L., Broad, S., Diekmann, D., Evans, R. D., and Watt, F. M. (2000) Mol. Biol. Cell
11, 453-466
44.
Bretscher, A., Edwards, K., and Fehon, R. G. (2002) Nat. Rev. Mol. Cell Biol. 3, 586-599
45.
Pearson, M. A., Reczek, D., Bretscher, A., and Karplus, P. A. (2000) Cell 101, 259-270
46.
Calderwood, D. A., Yan, B., de Pereda, J. M., Alvarez, B. G., Fujioka, Y., Liddington, R.
47.
Ulmer, T. S., Calderwood, D. A., Ginsberg, M. H., and Campbell, I. D. (2003)
Biochemistry 42, 8307-8312
48.
Tadokoro, S., Shattil, S. J., Eto, K., Tai, V., Liddington, R. C., de Pereda, J. M.,
Ginsberg, M. H., and Calderwood, D. A. (2003) Science 302, 103-106
49.
Girault, J. A., Labesse, G., Mornon, J. P., and Callebaut, I. (1999) Trends Biochem. Sci.
24, 54-57
50.
Zavadil, J., Bitzer, M., Liang, D., Yang, Y. C., Massimi, A., Kneitz, S., Piek, E., and
Bottinger, E. P. (2001) Proc. Natl. Acad. Sci. U. S. A. 98, 6686-6691
51.
Tu, Y., Wu, S., Shi, X., Chen, K., and Wu, C. (2003) Cell 113, 37-47
52.
Niessen, C. M., van der Raaij-Helmer, M. H., Hulsman, E. H., van der Neut, R.,
Jonkman, M. F., and Sonnenberg, A. (1996) J. Cell Sci. 109, 1695-1706
53.
Kreidberg, J. A. (2000) Curr. Opin. Cell Biol. 12, 548-553
54.
Georges-Labouesse, E., Messaddeq, N., Yehia, G., Cadalbert, L., Dierich, A., and Le
Meur, M. (1996) Nat. Genet. 13, 370-373
55.
van der Neut, R., Krimpenfort, P., Calafat, J., Niessen, C. M., and Sonnenberg, A. (1996)
Nat. Genet. 13, 366-369
28
Downloaded from www.jbc.org at University of North Carolina at Chapel Hill, on August 23, 2012
C., and Ginsberg, M. H. (2002) J. Biol. Chem. 277, 21749-21758
56.
Brakebusch, C., Grose, R., Quondamatteo, F., Ramirez, A., Jorcano, J. L., Pirro, A.,
Svensson, M., Herken, R., Sasaki, T., Timpl, R., Werner, S., and Fassler, R. (2000)
EMBO J. 19, 3990-4003
57.
Raghavan, S., Bauer, C., Mundschau, G., Li, Q., and Fuchs, E. (2000) J. Cell Biol. 150,
1149-1160
Haber, R. M., and Hanna, W. M. (1996) Arch. Dermatol. 132, 1487-1490
59.
Shimizu, H., Sato, M., Ban, M., Kitajima, Y., Ishizaki, S., Harada, T., BrucknerTuderman, L., Fine, J. D., Burgeson, R., Kon, A., McGrath, J. A., Christiano, A. M.,
Uitto, J., and Nishikawa, T. (1997) Arch. Dermatol. 133, 1111-1117
60.
DiPersio, C. M., Hodivala-Dilke, K. M., Jaenisch, R., Kreidberg, J. A., and Hynes, R. O.
(1997) J. Cell Biol. 137, 729-742
61.
Ignotz, R. A., and Massague, J. (1987) Cell 51, 189-197
62.
Koivisto, L., Larjava, K., Hakkinen, L., Uitto, V. J., Heino, J., and Larjava, H. (1999)
Cell Adhes. Commun. 7, 245-257
63.
Garbi, C., Colletta, G., Cirafici, A. M., Marchisio, P. C., and Nitsch, L. (1990) Eur. J.
Cell Biol. 53, 281-289
64.
Furuyama, A., Iwata, M., Hayashi, T., and Mochitate, K. (1999) Eur. J. Cell Biol. 78,
867-875
65.
Nguyen, N. M., Bai, Y., Mochitate, K., and Senior, R. M. (2002) Am. J. Physiol. Lung
Cell Mol. Physiol. 282, L1004-L1011
66.
Yi, J. Y., Hur, K. C., Lee, E., Jin, Y. J., Arteaga, C. L., and Son, Y. S. (2002) Eur. J. Cell
Biol. 81, 457-468
29
Downloaded from www.jbc.org at University of North Carolina at Chapel Hill, on August 23, 2012
58.
67.
Bellone, G., Carbone, A., Tibaudi, D., Mauri, F., Ferrero, I., Smirne, C., Suman, F.,
Rivetti, C., Migliaretti, G., Camandona, M., Palestro, G., Emanuelli, G., and Rodeck, U.
(2001) Eur. J. Cancer 37, 224-233
Downloaded from www.jbc.org at University of North Carolina at Chapel Hill, on August 23, 2012
30
FOOTNOTES
1
The abbreviations used are: TGFβ−1, transforming growth factor β; EMT, epithelial to
mesenchymal transition; HMEC, human mammary epithelial cells; Mig-2, mitogen inducible
gene-2; FERM, four point one ezrin radixin moesin; PH, pleckstrin homology; pat, paralyzed
arrested at two-fold; PTB, phosphotyrosine binding; CHO, Chinese hamster ovary cells; and
siRNA, small interfering siRNA.
Downloaded from www.jbc.org at University of North Carolina at Chapel Hill, on August 23, 2012
31
FIGURE LEGENDS
FIG 1. Comparison of the amino acid sequences of kindlerin paralogues and UNC-112.
Amino acid alignment of kindlerin, Mig-2, URP2, and UNC-112. Identical amino acids are
highlighted in black and dashes indicate gaps in the alignment. The amino acid sequences were
aligned using the Clustal method.
mRNA isolated from HMEC treated with vehicle (V) or TGFβ for the indicated time. B. HMEC
were treated with vehicle (-) or TGFβ (+) for three hours and total RNA was isolated. In parallel
cultures, cyclohexamide (CHX) was added to the cells 15 minutes prior to the addition of TGFβ.
Expression levels of kindlerin and GAPDH were analyzed by Northern blot.
FIG 3. Antibody characterization and kindlerin expression in TGF treated HMEC. A.
Western blot analysis of HMEC and HaCaT lysates (25 µg) using pre-immune (left panel) or
kindlerin antiserum (right panel). B. Time course analysis of HMEC treated with TGFβ.
Lysates (15 µg) were probed with kindlerin, vinculin, and E-cadherin antibodies.
C.
Quantification of kindlerin induction by TGFβ normalized to vinculin levels. D. Phase contrast
images of HMEC treated with vehicle (control) or TGFβ for 48 hours. Images were taken at the
same magnification.
FIG 4. Localization of kindlerin in mammalian cells. Indirect immunofluorescence analysis
of HaCaT cells using kindlerin (A) and vinculin (B) antibodies. Panel C is a merged image of A
and B. White arrows indicate sites of focal adhesions. Panel D shows combined staining of
32
Downloaded from www.jbc.org at University of North Carolina at Chapel Hill, on August 23, 2012
FIG 2. TGF treatment of HMEC. A. Northern blot analysis of kindlerin and GAPDH using
kindlerin (red) and F-actin (green). Yellow arrows point to kindlerin localization at the ends of
actin filaments. HaCaT cells stained with control (E) and vinculin (F) antibodies.
FIG 5. The effect of TGF treatment on kindlerin distribution and the actin cytoskeleton.
Staining patterns of kindlerin (red) and F-actin (green) in the absence (A-B) and presence (C-D)
of TGFβ in HaCaT cells. The images in panels A-D are shown at the same magnification and
FIG 6. The effect of TGF treatment on kindlerin distribution and vinculin localization.
Indirect immunofluorescence analysis of kindlerin (red) and vinculin (green) in the absence (AC) or presence (D-I) of TGFβ. The images in panels A-F are shown at the same magnification
and the bar indicates 50 µM. Panels G-I represent enlarged images of the boxed regions in
panels D-F. Yellow arrows indicate sites of kindlerin-enriched staining and white arrows point
to focal adhesions stained by vinculin antibodies.
FIG 7. Comparison of talin and kindlerin FERM domains.
A. A linear diagram of
kindlerin. The FERM domain is shown in green and blue to indicate the F2 and F3 subdomains,
respectively and the gray box indicates homology to talin before the predicted start of the F2
subdomain. The pleckstrin homology domain is indicated by a black box. Mutations identified
in Kindler syndrome by Jobard et al. (1) are indicated by asterisks at amino acids 128, 154, 262,
and 572. B. Alignment of kindlerin and talin FERM domains. The primary sequences of
kindlerin and chicken talin were aligned by
BLAST
searching and refined using structure based
alignments performed by Swiss-PbdViewer (http://us.expasy.org/spdbv/). Identical residues are
33
Downloaded from www.jbc.org at University of North Carolina at Chapel Hill, on August 23, 2012
the bar indicates 50 µM.
marked with * and similar residues with . below the alignment. The arrow indicates residue
572, site of one of the mutations identified in Kindler syndrome. The boxed residues in talin
mediate the β-integrin-talin interaction.
Kindlerin exhibits sequence similarity to the talin
FERM domain, and the talin FERM subdomains F2 and F3 are highlighted in green and blue,
respectively. Kindlerin contains a large insertion, residues 326-495, containing a predicted PH
C.
Model of the kindlerin F3 subdomain. Ribbon representations of the homology model of
the kindlerin F3 subdomain (left) and of the talin F3 subdomain structure with the integrin ligand
in green (PDB: 1MK7; right). Regions of kindlerin exhibiting primary sequence similarity to
talin are colored teal.
Modeling was performed and the figure generated using Swiss-
PdbViewer.
FIG 8. Integrin
tails form complexes with kindlerin. A. Cell lysate (150 µg) containing
FLAG-tagged kindlerin was mixed with beads coated with recombinant integrin cytoplasmic
tails or tails containing point mutations as indicated. Bound proteins were fractionated by SDSPAGE and binding of kindlerin assessed by Western blotting with anti-FLAG antibodies.
Loading of recombinant integrin tails onto the beads was assessed by Coomassie Blue staining.
B. Increasing amounts of cell lysate containing FLAG-tagged kindlerin were incubated with
integrin tails and analyzed as in panel A.
FIG 9. The effect of kindlerin knock-down on cell spreading. A. Western blot analysis of
lysate (20 µg) from control or kindlerin siRNA transfected HaCaT cells probed with kindlerin
antiserum (left panel). The membrane was stripped and probed with talin antibodies (right
34
Downloaded from www.jbc.org at University of North Carolina at Chapel Hill, on August 23, 2012
domain, within the putative FERM domain. These amino acids are not shown in the alignment.
panel) to demonstrate equivalent protein levels. B. Indirect immunofluorescence analysis of
control (left panel) or kindlerin (right panel) siRNA transfected HaCaT after plating on
fibronectin coated coverslips for 30 minutes using talin antiserum.
C.
The cell area of
approximately 200 cells was measured by recording images of talin immunofluorescence. The p
value is less than 0.0001 using a two-tailed unpaired t test. The data is representative of three
independent experiments.
Downloaded from www.jbc.org at University of North Carolina at Chapel Hill, on August 23, 2012
35
Human mammary epithelial cells were treated with TGFβ (10ng/ml) for three
hours. mRNA harvested from treated and non-treated HMEC samples (3
independent replicates) was reverse transcribed into cDNA in the presence of
either a green or red fluorescent tag. Competitive hybridization of the labeled
cDNAs on glass slides containing 4608 cDNAs spotted in duplicate revealed the
relative increase and decrease of specific mRNAs. Stringent data analysis
(greater than 1.5 fold in 4 of the 6 microarray ratios; p-value less than 0.05)
facilitated the identification of 17 TGFβ upregulated transcripts.
Gene Name
Fold Change
p-value1
Plasminogen activator inhibitor, type I (PAI1)
Integrin α2
Kindlerin
Transglutaminase
Transforming Growth Factor β-induced, 68kD
Connective Tissue Growth Factor (CTGF)
PMEPA1
Sox-4
Tubulin α-4 chain
Tis11D/ERF-2
Gadd45β
EST (accession# AI828370)
DNA G/T mismatch-binding protein
Thrombospondin 1
TGF-β inducible early protein (TIEG)
Early Growth Response (EGR-1)
Fibronectin
3.07
2.55
2.30
2.17
1.96
1.94
1.89
1.88
1.76
1.70
1.69
1.66
1.63
1.62
1.61
1.58
1.58
7.76E-07
4.46E-04
5.36E-05
5.07E-03
4.92E-03
4.32E-05
5.56E-03
1.66E-02
2.13E-02
3.22E-04
7.24E-04
6.81E-03
6.79E-05
1.68E-02
4.04E-05
3.63E-02
8.22E-05
1 p-values
were calculated using the Students T test from 6 independent
microarray expression ratios.
Downloaded from www.jbc.org at University of North Carolina at Chapel Hill, on August 23, 2012
Table 1: Genes Induced by TGFβ in HMEC
M------LSSTDFTFASWELVVRVDHPNEEQQKDVTLRVSGDLHVGGVMLKLVEQINISQ
MALDGIRMPDGCYADGTWELSVHVT----DLNRDITLRVTGEVHIGGVMLKLVEKLDVKK
MA--GMKTASGDYIDSSWELRVFVGEEDPEAES-VTLRVTGESHIGGVLLKIVEQINRKQ
MA---HLVEGTSIIDGKWQLPILVT----DLNIQRSISVLGNLNVGGLMLELVSECDVER
Kindlerin
Mig-2
URP2
UNC-112
55
57
58
54
DWSDFALWWEQKHCWLLKTHWTLDKYGVQADAKLLFTPQHKMLRLRLPNLKMVRLRVSFS
DWSDHALWWEKKRTWLLKTHWTLDKYGIQADAKLQFTPQHKLLRLQLPNMKYVKVKVNFS
DWSDHAIWWEQKRQWLLQTHWTLDKYGILADARLFFGPQHRPVILRLPNRRALRLRASFS
DWSDHALWWPEKRRWLQHTRSTLDQNGITAETQLEFTPMHKEARIQLPDMQMIDARVDFS
Kindlerin
Mig-2
URP2
UNC-112
115
117
118
114
AVVFKAVSDICKILNIRRSEELSLLK-PSGDYFKKKKKKDKN-----------------DRVFKAVSDICKTFNIRHPEELSLLK-KPRDPTKKKKKKLDD-----------------QPLFQAVAAICRLLSIRHPEELSLLR-APE---KKEKKKKEK-----------------VNSFKATKKLCRDLGIRYSEELSLKRYIPPEDLRRGTSDADNMNGPLSMRPGEESVGPMT
Kindlerin
Mig-2
URP2
UNC-112
156
158
156
174
--NKEPIIEDILNLE-----SSPTASGSSV---SPGL--------YSKTMTPI---YDPI
--QSE---DEALELE-----GPLITPGSGSIYSSPGL--------YSKTMTPT---YDAH
--EPE---EELYDLS-----KVVLAGG-----VAPAL--------F-------------LRKAAPIFASQSNLDMRRRGQSPALSQSGHIFNAHEMGTLPRHGTLPRGVSPSPGAYNDT
Kindlerin
Mig-2
URP2
UNC-112
195
197
179
234
NGTPASSTMTWFSDSPLTEQNCSILAFSQPPQSPEALADMYQPRSLVDKAKLNAGWLDSS
DGSPLSPTSAWFGDSALSEGNPGILAVSQPITSPEILAKMFKPQALLDKAKINQGWLDSS
RGMPAH-----FSDSAQTEACYHMLSRPQPPPDPLLLQRLPRPSSLSDKTQLHSRWLDSS
MRRTPIMPSISFSEGLENEQFDDALIHS-PRLAPSRDTPVFRPQNYVEKAAINRGWLDSS
Kindlerin
Mig-2
URP2
UNC-112
255
257
234
293
RSLMEQGIQEDEQLLLRFKYYSFFDLNPKYDAVRINQLYEQARWAILLEEIDCTEEEMLI
RSLMEQDVKENEALLLRFKYYSFFDLNPKYDAIRINQLYEQAKWAILLEEIECTEEEMMM
RCLMQQGIKAGDALWLRFKYYSFFDLDPKTDPVRLTQLYEQARWDLLLEEIDCTEEEMMV
RSLMEQGIFEGDIILLRFKFMNFFDLNPKYDPVRINQLYEQAKWSILLDEFDHTEEEATL
Kindlerin
Mig-2
URP2
UNC-112
315
317
294
353
FAALQYHISKLSLSAETQDFA-GESEVDEIEAALSNLEVTLEGGKADSLLEDITDIPKLA
FAALQYHINKLSIMTSENHLNNSDKEVDEVDAALSDLEITLEGGKTSTILGDITSIPELA
FAALQYHINKLSQSGEVGEPAGTDPGLDDLDVALSNLEVKLEGSAPTDVLDSLTTIPELK
FAALQ-----LQATLQRDSPEPEENNKDDVDILLDELEQNLDAAALNRR-SDLTQVPELA
Kindlerin
Mig-2
URP2
UNC-112
374
377
354
407
DNLKLFRPKKLL-PKAFKQYWFIFKDTSIAYFKNKELEQGEPLEKLNLRGCEVVPDVNVA
DYIKVFKPKKLT-LKGYKQYWCTFKDTSISCYKSKEESSGTPAHQMNLRGCEVTPDVNIS
DHLRIFRPRKLT-LKGYRQHWVVFKETTLSYYKSQDEAPGDPIQQLNLKGCEVVPDVNVS
DYLKYMKPKKLAAFKGFKRAFFSFRDLYLSYHQSSSDVNSAPLGHFSLKGCEVSQDVSVG
Kindlerin
Mig-2
URP2
UNC-112
433
436
413
467
GRKFGIKLLIPVADGMNEMYLRCDHENQYAQWMAACMLASKGKTMADSSYQPEVLNILSF
GQKFNIKLLIPVAEGMNEIWLRCDNEKQYAHWMAACRLASKGKTMADSSYNLEVQNILSF
GQKFCIKLLVPSPEGMSEIYLRCQDEQQYARWMAGCRLASKGRTMADSSYTSEVQAILAF
QQKYHIKLLLPTAEGMIDFILKCDSEHQYARWMAACRLASRGKSMADSSYQQEVESIKNL
Kindlerin
Mig-2
URP2
UNC-112
493
496
473
527
LRMK-----NRNSASQVA----SSLENMDMNPECFVSPRCAKRHKSKQ-LAARILEAHQN
LKMQ-----HLNPDPQLI----PEQITTDITPECLVSPRYLKKYKNKQ-ITARILEAHQN
LSLQ-----RTGSGGPGNHPHGPDASAEGLNPYGLVAPRFQRKFKAKQ-LTPRILEAHQN
LKMQSGNGNENGNSNTASRKAAAVKLPNDFNVDEYISSKYVRRARSKQQIQQRVSDAHGN
Kindlerin
Mig-2
URP2
UNC-112
543
546
527
587
VAQMPLVEAKLRFIQAWQSLPEFGLTYYLVRFKGSKKDDILGVSYNRLIKIDAATGIPVT
VAQMSLIEAKMRFIQAWQSLPEFGITHFIARFQGGKKEELIGIAYNRLIRMDASTGDAIK
VAQLSLAEAQLRFIQAWQSLPDFGISYVMVRFKGSRKDEILGIANNRLIRIDLAVGDVVK
VRQLTATEAKLQYIRAWQALPEHGIHYFIVRFRNARKAELVAVAVNRLAKLNMDNGESLK
Kindlerin
Mig-2
URP2
UNC-112
603
606
587
647
TWRFTNIKQWNVNWETRQVVIEFDQNVFTAFTCLSADCKIVHEYIGGYIFLSTRSKDQNE
TWRFSNMKQWNVNWEIKMVTVEFADEVRLSFICTEVDCKVVHEFIGGYIFLSTRAKDQNE
TWRFSNMRQWNVNWDIRQVAIEFDEHINVAFSCVSASCRIVHEYIGGYIFLSTRERARGE
TWRFANMKKWHVNWEIRHLKIQFEDE-DIEFKPLSADCKVVHEFIGGYIFLSMRSKEHSQ
Kindlerin
Mig-2
URP2
UNC-112
663
666
647
706
TLDEDLFHKLTGGQD
SLDEEMFYKLTSGWV
ELDEDLFLQLTGGHEAF
NLDEELFHKLTGGW-A
Kindlerin
Mig-2
URP2
UNC-112
Downloaded from www.jbc.org at University of North Carolina at Chapel Hill, on August 23, 2012
1
1
1
1
Figure 1
S. Kloeker, et al.
Downloaded from www.jbc.org at University of North Carolina at Chapel Hill, on August 23, 2012
A
hours post-TGFβ
V
2
4
6
8
10
12
Kindlerin
GAPDH
B
TGFβ:
CHX:
-
+
-
+
+
+
7.5
4.4
Kindlerin
2.4
1.3
GAPDH
Figure 2
S. Kloeker et al.
HM
E
Ha C
Ca
T
HM
E
Ha C
Ca
T
A
98
64
64
50
50
36
36
Hours post -TGFβ
V 4
8 12 24 48
Kindlerin
62
Downloaded from www.jbc.org at University of North Carolina at Chapel Hill, on August 23, 2012
98
B
188
Vinculin
98
98
E-cadherin
62
22
4
4
150
100
50
TGFβ
Net Intensity
Normalized to Vinculin (x103)
C
D
Control
22
0
V
4
8
12
24
48
Time (h)
Figure 3
S. Kloeker, et al.
Downloaded from www.jbc.org at University of North Carolina at Chapel Hill, on August 23, 2012
Figure 4
S. Kloeker, et al.
Downloaded from www.jbc.org at University of North Carolina at Chapel Hill, on August 23, 2012
Figure 5
S. Kloeker, et al.
Downloaded from www.jbc.org at University of North Carolina at Chapel Hill, on August 23, 2012
Figure 6
S. Kloeker, et al.
A
1
**
250
*
326
FERM
378
473
496
F2
653 677
*
PH
FERM
F2
F3
B
Kindlerin 250 WLDSSRSLME QGIQEDEQLL LRFKYYSFFD LN-PKYDAVR INQLYEQARW
Talin
173 WLDHGRTLRE QGIDDNETLL LRRKFF -YSD QNVDSRDPVQ LNLLYVQARD
*** .*.* * *** ..* ** ** *.. . * *
* *. .* ** ***
F2
Downloaded from www.jbc.org at University of North Carolina at Chapel Hill, on August 23, 2012
Kindlerin 299 AILLEEIDCT EEEMLIFAAL QYHISKL
Talin
222 DILNGSHPVS FDKACEFAGY QCQI -QF
**
. .
**. * .* .
residues 326-495 not shown
-------------------
Kindlerin 496 KNRNSASQVA SSLENMDMNP ECFVSPRCAK KHKSKQLAAR ILEAHQNVAQ
Talin
248 GPHNEQKHKP GFLELKDFLP KEYI ------ KQKGER---K IFMAHKNCGN
.*
. . . ** .* *
..
*.*. .
. *. * * * .
F2
Kindlerin 546 MPLVEAKLRF IQAWQSLPEF GLTYYLVR-- FKGS--KKDD ILGVSYNRLI
Talin
289 MSEIEAKVRY VKLARSLKTY GVSFFLVKEK MKGKNKLVPR LLGITKECVM
*. .***.*. .
.** . *....**.
**
.**.. . ..
F3
Kindlerin 592 KIDAATGIPV TTWRFTNIKQ WNVNWETRQV VIEFDQNVFT AFTCLSADCK
Talin
339 RVDEKTKEVI QEWSLTNIKR WAAS --PKSF TLDFGDYQDG YYSVQTTEGE
..* *
.
* .****. * .
.
..* .
.. ...
Kindlerin 642 IVHEYIGGYI FLSTRSKDQ
Talin
387 QIAQLIAGYI DIILKKKKS
. . *.*** .
. *
C
Kindlerin F3
Talin F3
Figure 7
S. Kloeker, et al.
A
Kindlerin
te
sa
Ly
A)
00
8
Ib
8,
αI
78
(Y
A
β1
A)
88
Y7
A(
β1
A
β1
te
sa
Ly
Ib
αI
)
9A
75
)
7A
74
(Y
β3
74
(Y
β3
β3
25 50 100 150 1 10
0 10
µg of lysate
loaded
µg of lysate in
binding assay
B
7,
Downloaded from www.jbc.org at University of North Carolina at Chapel Hill, on August 23, 2012
Integrin tail
Kindlerin
Integrin tail
Kindlerin
β3 tail loading
Figure 8
S. Kloeker, et al.
C
er
in
dl
Ki
n
on
tro
l
tro
l
dl
er
in
Ki
n
C
on
Downloaded from www.jbc.org at University of North Carolina at Chapel Hill, on August 23, 2012
A
188
97
Kindlerin 188
52
Talin
97
B
50 µM
Control siRNA
C
Kindlerin siRNA
Cell Area (µm2)
1,000
800
600
400
200
0
Control
siRNA
Kindlerin
siRNA
Figure 9
S. Kloeker, et al.